How robots could answer farm labour shortages

Millions of robots like livestock-monitoring drones and apple-harvesting machines could be used in agriculture in the next three to five years.

The scientific community is anticipating millions of robots will be used in agriculture across the world as a solution to critical farm labour shortages.

From AI-powered, livestock-monitoring drones to apple-harvesting robots, autonomous machinery and chemical spraying devices, technology is increasingly becoming an option for Australian farmers to improve productivity, efficiency and safety.

Monash University’s Department of Mechanical and Aerospace Engineering deputy head Associate Professor Chao Chen said he expected robot use to noticeably increase on Australian farms in the next three to five years.

“Farming is an economic pillar in Australia, but it is suffering from labour shortages,” Dr Chen said.

“Robotics are a primary solution which can allow farmers to optimise their resources, reduce their labour requirements and also reduce their environmental impact.

“Robotics can also increase farming efficiency and improve sustainable productivity. A lot of tasks can be automated now such as planting, weeding, harvesting, orchard canopy monitoring and many others.

Eugene Kok, Wesley Au, Chao Chen, Brad Fankhauser, Tianhao Liu, Hugh Zhou with the MARS (Monash Apple Retrieving System) robot. Picture: Zoe Phillips

“There is a common agreement in the scientific community that after five years, millions of robots will be used on farms worldwide.”

Last year the federal government’s Department of Employment and Workplace Relations estimated the number of advertised job vacancies relating to agriculture ranged from about 5000 to 10,000, while ABARES found about 57 per cent of Australian horticulture farms had difficulty recruiting workers during 2021–22.

As a result of the reduced labour availability, ABARES also found about 40 per cent of Australian horticulture farms used advanced machinery to reduce their demand for labour during the same financial year.

Dr Chen, who will be speaking at FutureAg Expo 2024, has been working on the development of an apple harvesting machine for the past six years alongside a team of PhD students and graduates.

He said, their MARS (Monash Apple Retrieving System) robot, which won Victorian Project of the Year at the 2023 Engineers Australia Excellence Awards, was almost ready for commercialisation.

“For orchards and orchard growers, developing this robot is the equivalent of landing on the moon in agriculture,” Dr Chen said.

“Apple harvesting is a very challenging task because there is an uncertain environment. This is a highly precise task and there is high interaction with the environment.

The robot has to identify the apples and be precise to grasp it, but not only find the target, but avoid obstacles like branches, twigs and leaves.

“With our development we started by scanning the orchard and designing AI technology to reconstruct the 3D environment,

“The robotics information can then command the robot to use the gripper to approach in the right direction and avoid obstacles then drop in the fruit in the bag.”

On a more broadacre scale, Drone-hand, an AI-driven livestock and property monitoring drone, is offering farmers an automated and efficient way of checking on their animals.

The Drone-hand in action. Picture: Supplied.

Founded by Edward Barraclough, an experienced drone handler as an aerial photographer and videographer, Drone-hand was inspired by his ageing father.

“I visited my father, who was 82 at the time and I was using a drone to take some photos over the property, when he said if it could check on his sheep it would mean he could stay on the farm for longer,” Mr Barraclough said.

“So firstly I needed to validate the problem and see if this was something that excited beyond my dad.

“I found it was a pain point and many farmers were time-poor, staff costs had increased and stock theft had also become an issue.

“From there it was about using a machinery learning AI model and image recognition model to locate, identify, count and classify livestock.

The view from Drone-hand. Picture: Supplied

“Particularly problems like deceased animals, those having issues and birthing.

“I think we will see a real adoption of automated technologies because of their efficiencies and the way they can save costs.

“As they become cheaper there will be more and more automation.”

Dr Chen said the commercialisation and pricing of farming robotics will dictate how widely Australian farmers adopt the technologies.

“Commercially, pricing is a critical factor. The prices of the robots must be comparable with what growers are spending,” he said.

“In my opinion if robots do not increase operation cost, farmers will be willing to accept it, but if the costs are much higher it is harder for farmers to bring them in.”